© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 4-LLAMASCOPE-RES-32K
    11. 29743
    Prev
    Next
    INDEX
    Explanations

    references to nuclear events and their consequences in Japan, particularly relating to radiation and safety measures post-disaster

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.4.hook_resid_post
    Hook Layer
    4
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ardım
    -0.17
    маз
    -0.16
     ngu
    -0.16
    Sdk
    -0.15
    SelectionMode
    -0.14
    KN
    -0.14
    anax
    -0.14
     ÙħÙĨت
    -0.14
     addCriterion
    -0.14
     Kann
    -0.14
    POSITIVE LOGITS
     Aw
    0.15
    ories
    0.15
    URI
    0.15
    dog
    0.14
    ocker
    0.14
    bis
    0.14
    pron
    0.14
    Unique
    0.14
    ively
    0.13
    -
    0.13
    Activations Density 0.209%

    No Known Activations